Conventional extraction of three-dimensional (3D) structure and/or position of an object may have limitations in that it is typically assumed that the object and any media around it presents simple and well behaved optical properties in the optical sensing of the object. If this limitation of well-behaved optical properties does not hold then conventional techniques will generally not produce precise results of a level that may be required in some applications. Similarly, if the sensing of 3D structure and/or position needs to be repeated for a plurality of similar objects, variation among the objects or variation among the media around the objects may cause further difficulties.
As such, there is a need for improved systems and methods of optical sensing through differing optical media.